Model Compression

Bitnet.cpp: Efficient Edge Inference for Ternary LLMs ACL 2025 BitStack: Any-Size Compression of Large Language Models in Variable Memory Environments ICLR 2025 Block Circulant Adapter for Large Language Models IJCAI 2025 BlockDialect: Block-wise Fine-grained Mixed Format Quantization for Energy-Efficient LLM Inference ICML 2025 BlockPruner: Fine-grained Pruning for Large Language Models ACL 2025 Bonsai: Gradient-free Graph Condensation for Node Classification ICLR 2025 Bridging Layout and RTL: Knowledge Distillation based Timing Prediction ICML 2025 CABS: Conflict-Aware and Balanced Sparsification for Enhancing Model Merging ICML 2025 Can Post-Training Quantization Benefit from an Additional QLoRA Integration? NAACL 2025 Can Students Beyond the Teacher? Distilling Knowledge from Teacher’s Bias AAAI 2025 CARVQ: Corrective Adaptor with Group Residual Vector Quantization for LLM Embedding Compression EMNLP 2025 CASP: Compression of Large Multimodal Models Based on Attention Sparsity CVPR 2025 Cassic: Towards Content-Adaptive State-Space Models for Learned Image Compression ICCV 2025 Catastrophic Failure of LLM Unlearning via Quantization ICLR 2025 CBQ: Cross-Block Quantization for Large Language Models ICLR 2025 CFSP: An Efficient Structured Pruning Framework for LLMs with Coarse-to-Fine Activation Information COLING 2025 Channel Merging: Preserving Specialization for Merged Experts AAAI 2025 CLaSp: In-Context Layer Skip for Self-Speculative Decoding ACL 2025 CLMTracing: Black-box User-level Watermarking for Code Language Model Tracing EMNLP 2025 ClusComp: A Simple Paradigm for Model Compression and Efficient Finetuning ACL 2025 ClusterAttn: KV Cache Compression under Intrinsic Attention Clustering ACL 2025 COAT: Compressing Optimizer states and Activations for Memory-Efficient FP8 Training ICLR 2025 Cobra: Extending Mamba to Multi-Modal Large Language Model for Efficient Inference AAAI 2025 CodecNeRF: Toward Fast Encoding and Decoding, Compact, and High-quality Novel-view Synthesis AAAI 2025 Code-Optimise: Self-Generated Preference Data for Correctness and Efficiency NAACL 2025